This folder contains the data and code needed to replicate all analyses
presented in "Beyond the Mean: How Thinking About The Distribution of Public
Opinion Reduces Politicians' Perceptual Errors" by Nicholas C. Dias, Jack Lucas,
and Lior Sheffer.

All code can be found in the script `replication_code.R`. Data can be found in
the `data` folder. Here is a description of each data file:

- bayes_model.RDS: Bayesian MRP model
- bayes_mrp_prop_error.RDS: Bayesian MRP estimates with propagated error
- cmb_2020_to_2023.RDS: Canadian Municipal Barometer survey data from 2020–2023
- cmb_avg.RDS: Canadian Municipal Barometer cross-survey averages
- issue_data.RDS: Policy-issue data for Figure 3 in Supplementary Materials
- mrp.RDS: MRP estimates by method and year
- token_order_data.RDS: Token-order data from perceived-distribution task
- token_value_data.RDS: Token-value data from perceived-distribution task

`codebook.xlsx` describes each data file. `perceived_dist_task.qsf` contains our
survey. `perceived_dist_task.js` contains the JavaScript code that we used
within Qualtrics to administer the perceived- distribution task.

This README file sits in the root directory. All file paths referenced in our
code files start from this directory. All analyses were executed with R version
4.3.2.
